Polar analyte analysis

[Uwe Neue]

Well, this is interesting...

From Google and Hans's info, I would conclude that NQ is not ionic at neutral pH. So again, no buffer needed to complicate things.

The fact that it did not elute in 90/10 is pretty much expected. Even on a diol column I would expect that you need somewhere around 25% water to elute the beast. Run a gradient from 10% to 50% water, and we will see where it shows up.

[mbicking]

If you want to go back to PFP, try using it in normal phase mode, at higher acetonitrile concentrations. I have worked on this phase a lot and under the right conditions you can get some interesting (although frustrating to understand) retention patterns.

Two possibilities:

1. Prepare a 0.01% acid solution (phosphoric or sulfuric would work fine) and try injecting at 50/50 acidic water/acn. Increase the acn to increase the retention. A further decrease in acid concentration may also increase retention, so you have two variables to work with. I am betting that with all those nitrogens, this molecule may pick up a positive charge at low pH. Cations are retained on PFP by both ionic and NP effects.

2. Prepare a 50 mM ammonium formate solution (buy the 10 M concentrate from Aldrich and dilute) and mix 50/50 with acn, and try the same experiment again. Increase acn or reduce buffer to increase retention. I have seen neutral bases retained well on PFP.

[bisnettrj2]

Thanks for all the replies. I'll try to answer the few questions that popped up overnight...

"I was wondering how much the 4.6x50-mm 2.6-um Kinetex costs. Thanks!" - $299, special promotion to get people to buy the new Kinetex technology.

I am running a 7.5 minute gradient to 50% water after a 2.5 minute hold at 90:10 ACN:H2O with a 1 mL/min flow rate at 30 degrees C. I'm then holding at 50:50 for 2.5 minutes, then ramping back to 90:10 and re-equilibrating for about ten minutes. The sample diluent was 90:10 ACN:H2O, and I injected 10 uL.

I'm going to set up a few different runs today (yes, this is how I spend my Saturdays) with different instrument conditions, where I'll vary the steepness of the gradient, and hold longer at 50:50. Would there be any chance of using 'too much' water? Can you go to 100% if necessary in this mode?

mbicking - thanks for the advice on the PFP. If I get a chance next week, and if the HILIC isn't working out, I'll try it.

[bisnettrj2]

Results so far do not seem promising...

Here's the first few minutes of my blank, which is 90:10 ACN:H2O. This is the full-scale portion of my baseline before the gradient portion of my method. This initial isocratic hold is at 90:10 ACN:H2O.




Here's first few minutes of the baseline of the first injection of my standard, which has been diluted in 90:10 ACN:H2O to match the initial column conditions.




And here's the full-scale image of my first injection of my standard. Note the very large peak.




It looks to me like the nitroguanidine is not being retained at all on the column, since the only difference between the two injections is the addition of nitroguanidine, and I'm not seeing a peak in the rest of the run, even after holding at 50:50 ACN:H2O for 10 minutes after my gradient from 90:10 to 50:50.

[Uwe Neue]

This was the diol column?
Was this the first gradient you ran, or had the column seen the 50/50 mix before?

You need a water layer on the surface to get retention. If it does not work on the diol, my suggestion would be to wait until you have the silica column in your hands. I can't see that you would have no retention for such a polar compound...

[bisnettrj2]

Yes, this is the diol column. Below is from their "HILIC Method Development Guidelines" pdf http://www.phenomenex.com/lib/gu71811109_l.pdf

"Luna® HILIC and Kinetex™ HILIC columns are shipped in HILIC mobile
phase (90 % acetonitrile/10 % 100 mM ammonium formate, pH 3.2) and
can be used immediately after equilibration with your HILIC mobile phase.
For best results we highly recommend that you equilibrate your HILIC
column with at least 20 column volumes of your mobile phase prior
to analyzing samples. In addition, maintaining at least 5 % water in your
mobile phase will greatly reduce the column equilibration time and thus
ensure stable retention times. If gradient elution is used, then care should
be taken to ensure complete column equilibration between injections –
failure to do so will result in retention time drift and poor reproducibility.
For best reproducibility, we recommend that you inject a blank sample
at the beginning of each sample set. Additionally when using an
autosampler, be sure to incorporate a needle wash step with acetonitrile"

I equilibrated the column with 90:10 Acetonitrile:Water for 20 column volumes, then ran two blank 10 uL injections of a 90:10 ACN:H2O mix, followed by the injection of a 0.2 ppm NQ standard in 90:10 ACN:H2O.

The only thing I don't have is a 'needle wash step' programmed into my autosampler, as I've never used one and don't know how to program it into my injection program. I'm not even really sure why I would need it, to be honest. BTW, I'm using an Agilent 1100 with the G1329A autosampler, G1312A binary pump, and the G1365B MWD detector.

The method is as follows:

1.0 mL/min, 10 uL injection, 30 degrees C column compartment, 270 nm detection wavelength.

A = Millipore Water, B = HPLC-grade Acetonitrile

Time | A:B
0 | 10:90
5 | 10:90
15 | 50:50
25 | 50:50
28 | 10:90
34 | 10:90

34 minute total run time.

Imagine how I feel if YOU'RE confused by this not working! Should I try to run a high percentage of water at some point to get a water layer on the surface, then re-equilibrate to 90:10 ACN:H2O?

Unfortunately, I am only getting this column, and no bare silica column is coming. I keep my eyes on eBay, though, in case one pops up that I can snag. It's hard to justify expenses on columns that *might* work in a production lab environment....

In any case, I made up about 10 different methods to try, wherein I will: vary the initial percentage of ACN to 95 and 97.5%; where I will be running a gradient to 100% water; where I will be using a 100 uL injection volume; and, in one method, where I will reduce the column flow rate to 0.1 mL/min from t0 to 0.2 minutes, then ramp the flow back to 1.0 mL/min by 1 minute into the run, in hopes that this will partition the analyte from the sample diluent and allow it to retain on-column (shot in the dark!). All these methods have 20-column-volume re-equilibration periods built into the back end of the run, and each method will shoot a blank, and then a NQ standard in 90:10 ACN:H2O and 100% H2O. I should have results by Monday.

Thanks for all the help, and I'll post Monday with results.

[Uwe Neue]

You can do one more thing. Your analyte will become more polar, it it is ionized. So you could go to pH 2 or so with phosphoric acid. The issue with phosphoric acid is the solubility in the mobile phase. I do not have anything here at home where I could look up the solubility of phosphoric acid in 90% acetonitrile. Maybe somebody else will know. If the analyte does not move, we are beating up a dead horse...

[bisnettrj2]

Beating a dead horse, my favorite activity. I think, if nothing that I did today works, I'll try 0.1% phosphoric acid in acetonitrile and in water, and I'll make my standards in the same.

Is there anything in HILIC theory that has to do with solubilities in the aprotic or organic solvent versus water? Solubilities are usually only listed in water; if something is soluble in water and acetonitrile almost equally, or so soluble that only in the most extreme circumstances would you see a preference for water over the other, then would it follow that HILIC wouldn't work for these analytes? That the similar solubility would result in the analyte diffusing into the mobile phase, rather than the stationary phase, if it was similarly soluble in either? I'm just throwing out an idea I had; I haven't read the research on HILIC and why it works.

[Vlad Orlovsky]

I have a method for something similar - cyanoguanidine in normal phase mode on Primesep 100 column. Our task was to separate this compounds from dinitrophenylacetic and benzoic acids. Retention of cyanoguanidine about 13 minutes.

Let me know if you are interested, I can send it to you.

[Hollow]

did you already found that publication?
Maybe a helpful guide for adaption to HPLC (unfortunately I don't have access to the complete article)

Separation and Determination of Nitroguanidine and Guanidine Nitrate by HP-TLC
Chromatographia, Volume 66, Numbers 3-4 / August 2007

http://www.springerlink.com/content/v5h112kp4u51l581/

[carls]

Based on the logP (~-1) this analyte should be retained in HILIC mode. I suggest adding 10mM ammonium acetate (no pH adjustment needed) to your 90/10 ACN/H2O mobile phase. Mobile phase salts can have interesting and not necessarily predictable results in HILIC. The salt not only affects ion exchange but also ion exclusion and possibly phase ratio.

[bisnettrj2]

So, I found an old article (from 1980!) about various explosives and some of their properties. NQ was listed, and there was a solubility chart.

http://www.scribd.com/doc/7758105/los-a ... ofile-data

At 20 C, NQ has the following solubility (grams per 100 g of solvent):

water - 0.36
Dimethyl sulfoxide - 24
Dimethyl formamide - 14
Methanol 0.3
Methyl ethyl ketone - 0.13
Butyl acetate - 0.07
n-octane - 0.003

Does that give anybody any ideas?

[Vlad Orlovsky]

bisnettrj2,

I emailed you file with the method. Please check you bulk/spam folder if you did not receive it.

[bisnettrj2]

Thanks Vlad, I did receive it. I really hope I can get this analysis to work with the columns I already have - the bosses are not happy with situations where I ask them if I can get something because I think it "might" work. I wish I could get a plain silica column to try out for this analysis without having to pay for it... but then, how many of us wish we had access to all column types at all times to do whatever we wished with them, right?

[Vlad Orlovsky]

We are offering free method development for our customer. You can send us a sample and we will run it i the lab. Our success rate is close to 100% and we usually do this in a day or two.

Also try any HILIC column in normal phase mode (not HILIC mobile phase, try heptane/ethanol)